Documentation for Kdump - The kexec-based Crash Dumping Solution

This document includes overview, setup and installation, and analysis information.

Overview

Kdump uses kexec to quickly boot to a dump-capture kernel whenever a dump of the system kernel’s memory needs to be taken (for example, when the system panics). The system kernel’s memory image is preserved across the reboot and is accessible to the dump-capture kernel.

You can use common commands, such as cp and scp, to copy the memory image to a dump file on the local disk, or across the network to a remote system.

Kdump and kexec are currently supported on the x86, x86_64, ppc64, ia64, s390x, arm and arm64 architectures.

When the system kernel boots, it reserves a small section of memory for the dump-capture kernel. This ensures that ongoing Direct Memory Access (DMA) from the system kernel does not corrupt the dump-capture kernel. The kexec -p command loads the dump-capture kernel into this reserved memory.

On x86 machines, the first 640 KB of physical memory is needed to boot, regardless of where the kernel loads. Therefore, kexec backs up this region just before rebooting into the dump-capture kernel.

Similarly on PPC64 machines first 32KB of physical memory is needed for booting regardless of where the kernel is loaded and to support 64K page size kexec backs up the first 64KB memory.

For s390x, when kdump is triggered, the crashkernel region is exchanged with the region [0, crashkernel region size] and then the kdump kernel runs in [0, crashkernel region size]. Therefore no relocatable kernel is needed for s390x.

All of the necessary information about the system kernel’s core image is encoded in the ELF format, and stored in a reserved area of memory before a crash. The physical address of the start of the ELF header is passed to the dump-capture kernel through the elfcorehdr= boot parameter. Optionally the size of the ELF header can also be passed when using the elfcorehdr=[size[KMG]@]offset[KMG] syntax.

With the dump-capture kernel, you can access the memory image through /proc/vmcore. This exports the dump as an ELF-format file that you can write out using file copy commands such as cp or scp. Further, you can use analysis tools such as the GNU Debugger (GDB) and the Crash tool to debug the dump file. This method ensures that the dump pages are correctly ordered.

Setup and Installation

Install kexec-tools

  1. Login as the root user.
  2. Download the kexec-tools user-space package from the following URL:

http://kernel.org/pub/linux/utils/kernel/kexec/kexec-tools.tar.gz

This is a symlink to the latest version.

The latest kexec-tools git tree is available at:

There is also a gitweb interface available at http://www.kernel.org/git/?p=utils/kernel/kexec/kexec-tools.git

More information about kexec-tools can be found at http://horms.net/projects/kexec/

  1. Unpack the tarball with the tar command, as follows:

    tar xvpzf kexec-tools.tar.gz
    
  2. Change to the kexec-tools directory, as follows:

    cd kexec-tools-VERSION
    
  3. Configure the package, as follows:

    ./configure
    
  4. Compile the package, as follows:

    make
    
  5. Install the package, as follows:

    make install
    

Build the system and dump-capture kernels

There are two possible methods of using Kdump.

  1. Build a separate custom dump-capture kernel for capturing the kernel core dump.
  2. Or use the system kernel binary itself as dump-capture kernel and there is no need to build a separate dump-capture kernel. This is possible only with the architectures which support a relocatable kernel. As of today, i386, x86_64, ppc64, ia64, arm and arm64 architectures support relocatable kernel.

Building a relocatable kernel is advantageous from the point of view that one does not have to build a second kernel for capturing the dump. But at the same time one might want to build a custom dump capture kernel suitable to his needs.

Following are the configuration setting required for system and dump-capture kernels for enabling kdump support.

System kernel config options

  1. Enable “kexec system call” in “Processor type and features.”:

    CONFIG_KEXEC=y
    
  2. Enable “sysfs file system support” in “Filesystem” -> “Pseudo filesystems.” This is usually enabled by default:

    CONFIG_SYSFS=y
    

    Note that “sysfs file system support” might not appear in the “Pseudo filesystems” menu if “Configure standard kernel features (for small systems)” is not enabled in “General Setup.” In this case, check the .config file itself to ensure that sysfs is turned on, as follows:

    grep 'CONFIG_SYSFS' .config
    
  3. Enable “Compile the kernel with debug info” in “Kernel hacking.”:

    CONFIG_DEBUG_INFO=Y
    

    This causes the kernel to be built with debug symbols. The dump analysis tools require a vmlinux with debug symbols in order to read and analyze a dump file.

Dump-capture kernel config options (Arch Independent)

  1. Enable “kernel crash dumps” support under “Processor type and features”:

    CONFIG_CRASH_DUMP=y
    
  2. Enable “/proc/vmcore support” under “Filesystems” -> “Pseudo filesystems”:

    CONFIG_PROC_VMCORE=y
    

    (CONFIG_PROC_VMCORE is set by default when CONFIG_CRASH_DUMP is selected.)

Dump-capture kernel config options (Arch Dependent, i386 and x86_64)

  1. On i386, enable high memory support under “Processor type and features”:

    CONFIG_HIGHMEM64G=y
    

    or:

    CONFIG_HIGHMEM4G
    
  2. On i386 and x86_64, disable symmetric multi-processing support under “Processor type and features”:

    CONFIG_SMP=n
    

    (If CONFIG_SMP=y, then specify maxcpus=1 on the kernel command line when loading the dump-capture kernel, see section “Load the Dump-capture Kernel”.)

  3. If one wants to build and use a relocatable kernel, Enable “Build a relocatable kernel” support under “Processor type and features”:

    CONFIG_RELOCATABLE=y
    
  4. Use a suitable value for “Physical address where the kernel is loaded” (under “Processor type and features”). This only appears when “kernel crash dumps” is enabled. A suitable value depends upon whether kernel is relocatable or not.

    If you are using a relocatable kernel use CONFIG_PHYSICAL_START=0x100000 This will compile the kernel for physical address 1MB, but given the fact kernel is relocatable, it can be run from any physical address hence kexec boot loader will load it in memory region reserved for dump-capture kernel.

    Otherwise it should be the start of memory region reserved for second kernel using boot parameter “crashkernel=Y@X”. Here X is start of memory region reserved for dump-capture kernel. Generally X is 16MB (0x1000000). So you can set CONFIG_PHYSICAL_START=0x1000000

  5. Make and install the kernel and its modules. DO NOT add this kernel to the boot loader configuration files.

Dump-capture kernel config options (Arch Dependent, ppc64)

  1. Enable “Build a kdump crash kernel” support under “Kernel” options:

    CONFIG_CRASH_DUMP=y
    
  2. Enable “Build a relocatable kernel” support:

    CONFIG_RELOCATABLE=y
    
Make and install the kernel and its modules.

Dump-capture kernel config options (Arch Dependent, ia64)

  • No specific options are required to create a dump-capture kernel for ia64, other than those specified in the arch independent section above. This means that it is possible to use the system kernel as a dump-capture kernel if desired.

    The crashkernel region can be automatically placed by the system kernel at run time. This is done by specifying the base address as 0, or omitting it all together:

    crashkernel=256M@0
    

    or:

    crashkernel=256M
    

    If the start address is specified, note that the start address of the kernel will be aligned to 64Mb, so if the start address is not then any space below the alignment point will be wasted.

Dump-capture kernel config options (Arch Dependent, arm)

  • To use a relocatable kernel, Enable “AUTO_ZRELADDR” support under “Boot” options:

    AUTO_ZRELADDR=y
    

Dump-capture kernel config options (Arch Dependent, arm64)

  • Please note that kvm of the dump-capture kernel will not be enabled on non-VHE systems even if it is configured. This is because the CPU will not be reset to EL2 on panic.

Extended crashkernel syntax

While the “crashkernel=size[@offset]” syntax is sufficient for most configurations, sometimes it’s handy to have the reserved memory dependent on the value of System RAM – that’s mostly for distributors that pre-setup the kernel command line to avoid a unbootable system after some memory has been removed from the machine.

The syntax is:

crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset]
range=start-[end]

For example:

crashkernel=512M-2G:64M,2G-:128M

This would mean:

  1. if the RAM is smaller than 512M, then don’t reserve anything (this is the “rescue” case)
  2. if the RAM size is between 512M and 2G (exclusive), then reserve 64M
  3. if the RAM size is larger than 2G, then reserve 128M

Boot into System Kernel

  1. Update the boot loader (such as grub, yaboot, or lilo) configuration files as necessary.

  2. Boot the system kernel with the boot parameter “crashkernel=Y@X”, where Y specifies how much memory to reserve for the dump-capture kernel and X specifies the beginning of this reserved memory. For example, “crashkernel=64M@16M” tells the system kernel to reserve 64 MB of memory starting at physical address 0x01000000 (16MB) for the dump-capture kernel.

    On x86 and x86_64, use “crashkernel=64M@16M”.

    On ppc64, use “crashkernel=128M@32M”.

    On ia64, 256M@256M is a generous value that typically works. The region may be automatically placed on ia64, see the dump-capture kernel config option notes above. If use sparse memory, the size should be rounded to GRANULE boundaries.

    On s390x, typically use “crashkernel=xxM”. The value of xx is dependent on the memory consumption of the kdump system. In general this is not dependent on the memory size of the production system.

    On arm, the use of “crashkernel=Y@X” is no longer necessary; the kernel will automatically locate the crash kernel image within the first 512MB of RAM if X is not given.

    On arm64, use “crashkernel=Y[@X]”. Note that the start address of the kernel, X if explicitly specified, must be aligned to 2MiB (0x200000).

Load the Dump-capture Kernel

After booting to the system kernel, dump-capture kernel needs to be loaded.

Based on the architecture and type of image (relocatable or not), one can choose to load the uncompressed vmlinux or compressed bzImage/vmlinuz of dump-capture kernel. Following is the summary.

For i386 and x86_64:

  • Use vmlinux if kernel is not relocatable.
  • Use bzImage/vmlinuz if kernel is relocatable.

For ppc64:

  • Use vmlinux

For ia64:

  • Use vmlinux or vmlinuz.gz

For s390x:

  • Use image or bzImage

For arm:

  • Use zImage

For arm64:

  • Use vmlinux or Image

If you are using an uncompressed vmlinux image then use following command to load dump-capture kernel:

kexec -p <dump-capture-kernel-vmlinux-image> \
--initrd=<initrd-for-dump-capture-kernel> --args-linux \
--append="root=<root-dev> <arch-specific-options>"

If you are using a compressed bzImage/vmlinuz, then use following command to load dump-capture kernel:

kexec -p <dump-capture-kernel-bzImage> \
--initrd=<initrd-for-dump-capture-kernel> \
--append="root=<root-dev> <arch-specific-options>"

If you are using a compressed zImage, then use following command to load dump-capture kernel:

kexec --type zImage -p <dump-capture-kernel-bzImage> \
--initrd=<initrd-for-dump-capture-kernel> \
--dtb=<dtb-for-dump-capture-kernel> \
--append="root=<root-dev> <arch-specific-options>"

If you are using an uncompressed Image, then use following command to load dump-capture kernel:

kexec -p <dump-capture-kernel-Image> \
--initrd=<initrd-for-dump-capture-kernel> \
--append="root=<root-dev> <arch-specific-options>"

Please note, that –args-linux does not need to be specified for ia64. It is planned to make this a no-op on that architecture, but for now it should be omitted

Following are the arch specific command line options to be used while loading dump-capture kernel.

For i386, x86_64 and ia64:

“1 irqpoll maxcpus=1 reset_devices”

For ppc64:

“1 maxcpus=1 noirqdistrib reset_devices”

For s390x:

“1 maxcpus=1 cgroup_disable=memory”

For arm:

“1 maxcpus=1 reset_devices”

For arm64:

“1 maxcpus=1 reset_devices”

Notes on loading the dump-capture kernel:

  • By default, the ELF headers are stored in ELF64 format to support systems with more than 4GB memory. On i386, kexec automatically checks if the physical RAM size exceeds the 4 GB limit and if not, uses ELF32. So, on non-PAE systems, ELF32 is always used.

    The –elf32-core-headers option can be used to force the generation of ELF32 headers. This is necessary because GDB currently cannot open vmcore files with ELF64 headers on 32-bit systems.

  • The “irqpoll” boot parameter reduces driver initialization failures due to shared interrupts in the dump-capture kernel.

  • You must specify <root-dev> in the format corresponding to the root device name in the output of mount command.

  • Boot parameter “1” boots the dump-capture kernel into single-user mode without networking. If you want networking, use “3”.

  • We generally don’t have to bring up a SMP kernel just to capture the dump. Hence generally it is useful either to build a UP dump-capture kernel or specify maxcpus=1 option while loading dump-capture kernel. Note, though maxcpus always works, you had better replace it with nr_cpus to save memory if supported by the current ARCH, such as x86.

  • You should enable multi-cpu support in dump-capture kernel if you intend to use multi-thread programs with it, such as parallel dump feature of makedumpfile. Otherwise, the multi-thread program may have a great performance degradation. To enable multi-cpu support, you should bring up an SMP dump-capture kernel and specify maxcpus/nr_cpus, disable_cpu_apicid=[X] options while loading it.

  • For s390x there are two kdump modes: If a ELF header is specified with the elfcorehdr= kernel parameter, it is used by the kdump kernel as it is done on all other architectures. If no elfcorehdr= kernel parameter is specified, the s390x kdump kernel dynamically creates the header. The second mode has the advantage that for CPU and memory hotplug, kdump has not to be reloaded with kexec_load().

  • For s390x systems with many attached devices the “cio_ignore” kernel parameter should be used for the kdump kernel in order to prevent allocation of kernel memory for devices that are not relevant for kdump. The same applies to systems that use SCSI/FCP devices. In that case the “allow_lun_scan” zfcp module parameter should be set to zero before setting FCP devices online.

Kernel Panic

After successfully loading the dump-capture kernel as previously described, the system will reboot into the dump-capture kernel if a system crash is triggered. Trigger points are located in panic(), die(), die_nmi() and in the sysrq handler (ALT-SysRq-c).

The following conditions will execute a crash trigger point:

If a hard lockup is detected and “NMI watchdog” is configured, the system will boot into the dump-capture kernel ( die_nmi() ).

If die() is called, and it happens to be a thread with pid 0 or 1, or die() is called inside interrupt context or die() is called and panic_on_oops is set, the system will boot into the dump-capture kernel.

On powerpc systems when a soft-reset is generated, die() is called by all cpus and the system will boot into the dump-capture kernel.

For testing purposes, you can trigger a crash by using “ALT-SysRq-c”, “echo c > /proc/sysrq-trigger” or write a module to force the panic.

Write Out the Dump File

After the dump-capture kernel is booted, write out the dump file with the following command:

cp /proc/vmcore <dump-file>

Analysis

Before analyzing the dump image, you should reboot into a stable kernel.

You can do limited analysis using GDB on the dump file copied out of /proc/vmcore. Use the debug vmlinux built with -g and run the following command:

gdb vmlinux <dump-file>

Stack trace for the task on processor 0, register display, and memory display work fine.

Note: GDB cannot analyze core files generated in ELF64 format for x86. On systems with a maximum of 4GB of memory, you can generate ELF32-format headers using the –elf32-core-headers kernel option on the dump kernel.

You can also use the Crash utility to analyze dump files in Kdump format. Crash is available on Dave Anderson’s site at the following URL:

Trigger Kdump on WARN()

The kernel parameter, panic_on_warn, calls panic() in all WARN() paths. This will cause a kdump to occur at the panic() call. In cases where a user wants to specify this during runtime, /proc/sys/kernel/panic_on_warn can be set to 1 to achieve the same behaviour.

Contact

GDB macros

#
# This file contains a few gdb macros (user defined commands) to extract
# useful information from kernel crashdump (kdump) like stack traces of
# all the processes or a particular process and trapinfo.
#
# These macros can be used by copying this file in .gdbinit (put in home
# directory or current directory) or by invoking gdb command with
# --command=<command-file-name> option
#
# Credits:
# Alexander Nyberg <alexn@telia.com>
# V Srivatsa <vatsa@in.ibm.com>
# Maneesh Soni <maneesh@in.ibm.com>
#

define bttnobp
        set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)
        set $pid_off=((size_t)&((struct task_struct *)0)->thread_group.next)
        set $init_t=&init_task
        set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)
        set var $stacksize = sizeof(union thread_union)
        while ($next_t != $init_t)
                set $next_t=(struct task_struct *)$next_t
                printf "\npid %d; comm %s:\n", $next_t.pid, $next_t.comm
                printf "===================\n"
                set var $stackp = $next_t.thread.sp
                set var $stack_top = ($stackp & ~($stacksize - 1)) + $stacksize

                while ($stackp < $stack_top)
                        if (*($stackp) > _stext && *($stackp) < _sinittext)
                                info symbol *($stackp)
                        end
                        set $stackp += 4
                end
                set $next_th=(((char *)$next_t->thread_group.next) - $pid_off)
                while ($next_th != $next_t)
                        set $next_th=(struct task_struct *)$next_th
                        printf "\npid %d; comm %s:\n", $next_t.pid, $next_t.comm
                        printf "===================\n"
                        set var $stackp = $next_t.thread.sp
                        set var $stack_top = ($stackp & ~($stacksize - 1)) + stacksize

                        while ($stackp < $stack_top)
                                if (*($stackp) > _stext && *($stackp) < _sinittext)
                                        info symbol *($stackp)
                                end
                                set $stackp += 4
                        end
                        set $next_th=(((char *)$next_th->thread_group.next) - $pid_off)
                end
                set $next_t=(char *)($next_t->tasks.next) - $tasks_off
        end
end
document bttnobp
        dump all thread stack traces on a kernel compiled with !CONFIG_FRAME_POINTER
end

define btthreadstack
        set var $pid_task = $arg0

        printf "\npid %d; comm %s:\n", $pid_task.pid, $pid_task.comm
        printf "task struct: "
        print $pid_task
        printf "===================\n"
        set var $stackp = $pid_task.thread.sp
        set var $stacksize = sizeof(union thread_union)
        set var $stack_top = ($stackp & ~($stacksize - 1)) + $stacksize
        set var $stack_bot = ($stackp & ~($stacksize - 1))

        set $stackp = *((unsigned long *) $stackp)
        while (($stackp < $stack_top) && ($stackp > $stack_bot))
                set var $addr = *(((unsigned long *) $stackp) + 1)
                info symbol $addr
                set $stackp = *((unsigned long *) $stackp)
        end
end
document btthreadstack
         dump a thread stack using the given task structure pointer
end


define btt
        set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)
        set $pid_off=((size_t)&((struct task_struct *)0)->thread_group.next)
        set $init_t=&init_task
        set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)
        while ($next_t != $init_t)
                set $next_t=(struct task_struct *)$next_t
                btthreadstack $next_t

                set $next_th=(((char *)$next_t->thread_group.next) - $pid_off)
                while ($next_th != $next_t)
                        set $next_th=(struct task_struct *)$next_th
                        btthreadstack $next_th
                        set $next_th=(((char *)$next_th->thread_group.next) - $pid_off)
                end
                set $next_t=(char *)($next_t->tasks.next) - $tasks_off
        end
end
document btt
        dump all thread stack traces on a kernel compiled with CONFIG_FRAME_POINTER
end

define btpid
        set var $pid = $arg0
        set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)
        set $pid_off=((size_t)&((struct task_struct *)0)->thread_group.next)
        set $init_t=&init_task
        set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)
        set var $pid_task = 0

        while ($next_t != $init_t)
                set $next_t=(struct task_struct *)$next_t

                if ($next_t.pid == $pid)
                        set $pid_task = $next_t
                end

                set $next_th=(((char *)$next_t->thread_group.next) - $pid_off)
                while ($next_th != $next_t)
                        set $next_th=(struct task_struct *)$next_th
                        if ($next_th.pid == $pid)
                                set $pid_task = $next_th
                        end
                        set $next_th=(((char *)$next_th->thread_group.next) - $pid_off)
                end
                set $next_t=(char *)($next_t->tasks.next) - $tasks_off
        end

        btthreadstack $pid_task
end
document btpid
        backtrace of pid
end


define trapinfo
        set var $pid = $arg0
        set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)
        set $pid_off=((size_t)&((struct task_struct *)0)->thread_group.next)
        set $init_t=&init_task
        set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)
        set var $pid_task = 0

        while ($next_t != $init_t)
                set $next_t=(struct task_struct *)$next_t

                if ($next_t.pid == $pid)
                        set $pid_task = $next_t
                end

                set $next_th=(((char *)$next_t->thread_group.next) - $pid_off)
                while ($next_th != $next_t)
                        set $next_th=(struct task_struct *)$next_th
                        if ($next_th.pid == $pid)
                                set $pid_task = $next_th
                        end
                        set $next_th=(((char *)$next_th->thread_group.next) - $pid_off)
                end
                set $next_t=(char *)($next_t->tasks.next) - $tasks_off
        end

        printf "Trapno %ld, cr2 0x%lx, error_code %ld\n", $pid_task.thread.trap_no, \
                                $pid_task.thread.cr2, $pid_task.thread.error_code

end
document trapinfo
        Run info threads and lookup pid of thread #1
        'trapinfo <pid>' will tell you by which trap & possibly
        address the kernel panicked.
end

define dump_log_idx
        set $idx = $arg0
        if ($argc > 1)
                set $prev_flags = $arg1
        else
                set $prev_flags = 0
        end
        set $msg = ((struct printk_log *) (log_buf + $idx))
        set $prefix = 1
        set $newline = 1
        set $log = log_buf + $idx + sizeof(*$msg)

        # prev & LOG_CONT && !(msg->flags & LOG_PREIX)
        if (($prev_flags & 8) && !($msg->flags & 4))
                set $prefix = 0
        end

        # msg->flags & LOG_CONT
        if ($msg->flags & 8)
                # (prev & LOG_CONT && !(prev & LOG_NEWLINE))
                if (($prev_flags & 8) && !($prev_flags & 2))
                        set $prefix = 0
                end
                # (!(msg->flags & LOG_NEWLINE))
                if (!($msg->flags & 2))
                        set $newline = 0
                end
        end

        if ($prefix)
                printf "[%5lu.%06lu] ", $msg->ts_nsec / 1000000000, $msg->ts_nsec % 1000000000
        end
        if ($msg->text_len != 0)
                eval "printf \"%%%d.%ds\", $log", $msg->text_len, $msg->text_len
        end
        if ($newline)
                printf "\n"
        end
        if ($msg->dict_len > 0)
                set $dict = $log + $msg->text_len
                set $idx = 0
                set $line = 1
                while ($idx < $msg->dict_len)
                        if ($line)
                                printf " "
                                set $line = 0
                        end
                        set $c = $dict[$idx]
                        if ($c == '\0')
                                printf "\n"
                                set $line = 1
                        else
                                if ($c < ' ' || $c >= 127 || $c == '\\')
                                        printf "\\x%02x", $c
                                else
                                        printf "%c", $c
                                end
                        end
                        set $idx = $idx + 1
                end
                printf "\n"
        end
end
document dump_log_idx
        Dump a single log given its index in the log buffer.  The first
        parameter is the index into log_buf, the second is optional and
        specified the previous log buffer's flags, used for properly
        formatting continued lines.
end

define dmesg
        set $i = log_first_idx
        set $end_idx = log_first_idx
        set $prev_flags = 0

        while (1)
                set $msg = ((struct printk_log *) (log_buf + $i))
                if ($msg->len == 0)
                        set $i = 0
                else
                        dump_log_idx $i $prev_flags
                        set $i = $i + $msg->len
                        set $prev_flags = $msg->flags
                end
                if ($i == $end_idx)
                        loop_break
                end
        end
end
document dmesg
        print the kernel ring buffer
end